Using virtual metrics for assessing the quality of color images

Author(s) Collection number Pages Download abstract Download full text
Ткаченко В. П., Гордєєв А. С. № 1 (85) 65-72 Image Image

The problem of using virtual metrics to assess the quality of color images is considered. When working with images, the problem of assessing the quality of a graphic object arises. Currently, the most common methods for assessing quality are comparing the current image with a reference sample. The degree of closeness, and therefore the degree of quality, can be, for example, the standard deviation or the correlation coefficient. Despite the simplicity and efficiency of this approach, in practice it is not always possible to use it due to the lack of a reference image. Quality control of a compressed image is a complex and controversial process. Traditional image property analysis features include PSNR and MSE metrics, as well as structural features (SSIM) and multi-scale MSSIM. The optimal values of the metric indicators corresponding to the images with the best sharpness are determined – the same as in the case of visual expert evaluation. The considered metrics are based on the assessment of image sharpness. In the course of experiments and analysis of more than 100 JPEG-compressed photorealistic images with different detail, the standard deviation MFSDmin ≈ 0.5 is obtained, at which distortions become hardly noticeable to the eye. The proposed approach is advisable to use for a preliminary assessment of the quality of the resulting image at the stage of preparation for printing. The PSNR metrics contain characteristics that are estimated based on the contrast sensitivity properties of vision. During subsequent printing of experimental samples, the boundaries of broken lines are smoothed compared to the original file: on the one hand, when printing, the width of the lines increases (in the direction of printing), on the other hand, ink leakage smooths out the roughness of the line boundaries.

Keywords: visual metrics, quality, contrast, difference, color images, decomposition.

doi: 10.32403/0554-4866-2023-1-85-65-72

  • 1. Pérez-Rodríguez, Fernando, & Gómez-García, Esteban. (2019). Codelplant: Regression-based processing of RGB images for color models in plant image segmentation: Computers and Electronics in Agriculture, 163 (in English).
  • 2. Sunyoung, Cho, & Hyeran, Byun. (2022). Dynamic curve color model for image matting: Pattern Recognition Letters, 33, 920–933 (in English).
  • 3. Chunzhi, Gu, Xuequan, Lu, & Chao, Zhang. (2022). Example-based color transfer with Gaussian mixture modeling: Pattern Recognition, 129 (in English).
  • 4. Peng, Lu, Xujun, Peng, & Xiaojie, Wang. (2016). Image color harmony modeling through neighbored co-occurrence colors: Neurocomputing, 201, 82–91 (in English).
  • 5. Schloss, Karen B., Lessard, Laurent, & Hurlbert, Anya C. (2017). Modeling color preference using color space metrics: Vision Research, 151, 99–116 (in English).
  • 6. Pulla, Victor, & Serrano, Xavier. (2015). Modeling of a neuro fuzzy system to develop an efficient method to get a specific color paint from the color model cyan, magenta and yellow (CMY) under terms of open source: Procedia Computer Science, 61, 486–491 (in English).
  • 7. Bereg, Sergey, Ma, Feifei, & Zhu, Binhai. (2018). On some matching problems under the color-spanning model: Theoretical Computer Science, 786, 26–31 (in English).
  • 8. Peng, Lu, Xujun, Peng, & Xiaojie, Wang. (2015). Towards aesthetics of image: A Bayesian framework for color harmony modeling: Signal Processing: Image Communication, 39, 487–498 (in English).
  • 9. Leyuan, Wu, Xiaogang, Zhang, & Yicong, Zhou. (2020). Unsupervised quaternion model for blind colour image quality assessment: Signal Processing, 176 (in English).
  • 10. Kovalskyi, B. M., Dudiak, V. O., Zanko, N. V., & Pysanchyn, N. S. (2018). Vzaiemozv’iazok osnovopolozhnykh poniat teorii koloru z kolorovidtvorenniam u suchasnykh tsyfrovykh sys­temakh: Polihrafiia i vydavnycha sprava, 1 (75), 19–30 (in Ukrainian).
  • 11. Afanasiev, D. V., & Zorenko, Ya. V. (2019). Systematyzatsiia tekhnolohii stysnennia zobrazhen u system polihrafichnoho reprodukuvannia: Tekhnolohiia i tekhnika drukarstva, 1 (63), 45–57 (in Ukrainian).
  • 12. Kovalskyi, B. M. (2020). Informatsiina tekhnolohiia koloropodilu zobrazhennia. Lviv : Ukr. akad. drukarstva, 36−58 (in Ukrainian).
  • 13. Petrova, K., Zelenyi, O., & Deineko, Zh. (2022). Peredumovy vrakhuvannia kolirnoho pros­to­ru pry vybori tekhnolohii dlia obrobky zobrazhen. Polihrafichni, multymediini ta web-tekhnolohii. KhNURE, 74–75 (in Ukrainian).